statistics Flashcards
median
middle value
odd = (n+1)/2
even = (n/2+(n+1)/2)/2
mode
most frequent value
absolute frequency
count/number
relative frequency
percent of observations
cumulative absolute/relative frequency
skumulowane wartości procenty
outliers
wartości graniczne
delete = trimmed mean
replace = winsorised mean
quartile/
quintiles/
decilies/
percentiles
kwartyle/
n/5
n/10
n/100
co n-ty wyraz
disperation
rozproszenie
central tendency
tendencja centralna
range
max vol - miz vol
when no shape of the distribution, sensitive to outliers
mean absolute deviation
suma|x-xsr|/n
sample variance/popul variance
sample = mianowki (n-1)
popul = mianownik = n
downside deviation/semideviation
to samo tylko zamiast od średniej odejmujemy od jakiejś minimalnej wartości i dzielimy przez n-1. ważne że interesują nas jedynie mniejsze wartości
when we want a return x% instead of mean we use x% because we are concerned about values below this level
coefficient of variation (CV)
CV=S/xśr, measure of relative dispertion (procentowo). level of risk per unit of return (odchylenie standardowe/średnia)
the lower the less risk
Skewness
positively = more on the begining
negatively = more at the end
interquartile range
the difference between the third and first quartiles
the higher the dispersion of a distribution, the greater the difference between the arithmetic mean
kurtosis
k = 0 - normal distribution
k>3 - higher at the head thinner at the shoulders more on tail. = fat-tailed (leptokurtic)
k<3 lower at the head more at the shoulders lower tail. = thin-tailed (platokurtic)
excess kurtoisis
kurtoisis - 3
limitations of correlation
- Spurious correlation (pozorna korelacja)
- linear association only
- correlation does not imply causation
- not good with outliers
n-tiles number
The position of the first quintile is found with the following formula:
Ly = (n + 1) × (y/100),where
y = the percentage point at which the distribution is divided. In this case, y = 20, which corresponds to the 20th percentile (first quintile)
n = the number of observations (funds) in the peer group. In this case, n = 13